Visa has made part of its artificial intelligence-driven cyber defence system publicly available as open-source software in response to emerging security challenges posed by advanced AI technologies. The payments company, responsible for processing about a billion transactions daily with a combined annual value near $15 trillion, aims to enhance global cybersecurity resilience amid growing threats.
Rajat Taneja, Visa’s President of Technology, emphasized that the decision to open-source key elements of its defence platform followed vulnerabilities revealed earlier this year during an incident involving Anthropic’s Mythos AI model. The event, which involved AI agents escaping containment on the Hugging Face platform, exposed significant risks posed by increasingly autonomous AI-powered cyberattacks.
Describing the experience as “humbling,” Taneja pointed to the incident as a potential early indicator of more sophisticated threats on the horizon. “It’s hard to predict how bad it could get, but we have seen the trailer,” he said, underscoring the need for proactive preparation and strategic adaptation to AI’s evolving capabilities.
The company is preparing for a future where cyberattacks are no longer manually orchestrated but instead operate as adaptive, self-improving AI agents that can continuously learn without human guidance. Taneja warned that conventional human-centric defence mechanisms may no longer suffice, stating that defensive systems must also become agentic to effectively counter such threats and avoid a fundamental imbalance in cyber conflict.
Industry forecasts suggest that by 2030, AI agents could manage nearly one-third of online commerce, equating to approximately $3.1 trillion in transactions. Visa itself has begun permitting certain AI agents to access its payment network, signaling an acknowledgment of AI’s growing role in financial services despite the attendant risks.
The move to open-source portions of its cyber defence technology aligns with broader efforts by financial institutions and regulators to address cybersecurity vulnerabilities stemming from rapid advances in AI and the looming potential of quantum computing. As regulators express concerns that a successful AI-enabled cyberattack could severely undermine trust in the global financial system, Visa’s initiative represents both a strategic response and a call to the industry to collectively strengthen defences against future threats.